In October 2018, we disclosed the first comprehensive archive of state-backed information operations on Twitter. We launched this initiative to empower academic and public understanding of these coordinated campaigns around the world, and to empower independent, third-party scrutiny of these tactics on our platform.
As our investigations of platform manipulation around the world have continued, we subsequently added new datasets in January, June and August, while also sharing insights on Twitter’s internal investigative approach and how these complex, sometimes cross-jurisdictional operations are identified.
Nearly one year on, the archive is now the largest of its kind in the industry. Thousands of researchers have made use of these datasets that contain millions of individual Tweets and more than one terabyte of media. Using our archive, these researchers have conducted their own investigations and shared their insights and independent analyses with the world.
Transparency and openness are deep-seated values at the heart of Twitter which define and guide our methodology around these disclosures. Today, we are disclosing six additional datasets covering five separate jurisdictions. Per our policies on platform manipulation, we have permanently suspended all the below accounts from the service.
United Arab Emirates & Egypt
We have removed a network of 271 accounts originating in the United Arab Emirates (UAE) and Egypt. These accounts were interconnected in their goals and tactics: a multi-faceted information operation primarily targeting Qatar, and other countries such as Iran. It also amplified messaging supportive of the Saudi government.
We also found evidence that these accounts were created and managed by DotDev, a private technology company operating in the UAE and Egypt. We have permanently suspended DotDev, and all accounts associated with them, from our service.
Additionally, we suspended a separate group of 4,248 accounts operating uniquely from the UAE, mainly directed at Qatar and Yemen. These accounts were often employing false personae and tweeting about regional issues, such as the Yemeni Civil War and the Houthi Movement.
Our investigations also detected a small group of six accounts linked to Saudi Arabia’s state-run media apparatus which were engaged in coordinated efforts to amplify messaging that was beneficial to the Saudi government. While active, the accounts in this set presented themselves as independent journalistic outlets while tweeting narratives favourable to the Saudi government.
Separately, we have also permanently suspended the Twitter account of Saud al-Qahtani for violations of our platform manipulation policies. This account is not included in the archives disclosed today.
We continue to suspend a range of accounts for other types of political spam violations in Saudi Arabia. As ever, we will disclose any attribution we can reliably make to state actors if and when we learn them.
We have removed 259 accounts we identified as falsely boosting public sentiment online in Spain. Operated by Partido Popular, these accounts were active for a relatively short period, and consisted primarily of fake accounts engaging in spamming or retweet behaviour to increase engagement.
Earlier in the summer, we removed a network of 1,019 accounts in Ecuador tied to the PAIS Alliance political party. The network, composed largely of fake accounts, was primarily engaged in spreading content about President Moreno’s administration, focusing on issues concerning Ecuadorian laws on freedom of speech, government censorship, and technology. Tactics most commonly used were hashtag manipulation and retweet spam.
China (PRC)/Hong Kong
In August, we disclosed that we had identified a network of more than 200,000 fake accounts based in the PRC which were attempting to sow discord about the protest movement in Hong Kong. Today, we are publishing additional datasets relating to 4,301 accounts which were most active in this information operation to further public awareness and understanding.
Going forward, we will continue to enhance and refine our approach to disclosing state-affiliated information operations on our service. For example, we intend to more routinely disclose data relating to state-backed information operations on the service using our @TwitterSafety handle and in future iterations of the Twitter Transparency Report. We will share more on these plans soon.
We see this as a necessary and vital step to regularize our transparency work in this area, and to encourage more third-party investigation of these behaviors online.
You can access relevant information on these account sets below.
Did someone say … cookies?